Rogue A.I. Agents from China Chatted with Researchers (nytimes.com)

Researchers discovered that the agents were trying to bypass security controls on websites and made contact with them. A Times reporter was invited to join the discussion.